Real-Time Streaming Content Delivery via Cohort-Based Manifest Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing solutions for inserting advertisements in live streaming media are inefficient and unable to scale to large numbers of concurrent users, leading to high processing times and delays in advertisement insertion.
Innovation Solution
A method and system for real-time delivery of target content in streaming media, which involves receiving user segment information, dynamically fetching target content, creating a customized virtual manifest, and delivering the target content in real-time to user devices based on the manifest.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If personalized ad insertion is done on a per user basis for millions of concurrent users, then advertisement customization per user interest is improved, but processing time becomes very high and system scalability deteriorates
Solution Approach 1:
The patent segments users into cohorts based on common characteristics (location, device type, network) rather than processing each user individually. This segmentation allows the system to generate ad inserts once per cohort and serve them to multiple users simultaneously, reducing processing time from O(n) to O(1) per user while maintaining personalization through cohort-level customization.
Solution Approach 2:
The patent merges multiple per-user ad insertion operations into a single cohort-based ad insertion operation. By combining users with similar characteristics into cohorts, the system processes ad insertion once for the entire cohort rather than separately for each user, significantly improving processing throughput while maintaining personalized advertising through cohort-specific ad selection.
2Adaptability or versatility
If ad insertion processing is done separately for each user, then user-specific advertisement delivery is improved, but system scalability to millions of users deteriorates
Solution Approach 1:
The patent implements segmentation by dividing the large user base into manageable cohorts based on shared characteristics. This segmentation reduces the system complexity from handling millions of individual users to managing a smaller number of cohorts, making the system scalable while preserving user-specific ad delivery through cohort-level personalization.
Solution Approach 2:
The patent uses copying by generating ad insertion manifests once per cohort and then replicating them for all users within that cohort. This copying approach eliminates the need to regenerate ad inserts for each individual user, reducing system complexity and improving scalability while maintaining consistent personalized advertising across the cohort.
3Adaptability or versatility
If personalized advertisements are served to millions of users in real-time, then advertisement relevance to user interests is improved, but delivery delay increases due to high processing time
Solution Approach 1:
The patent applies preliminary action by pre-computing ad insertion manifests for cohorts before user requests arrive. The system proactively generates and caches ad inserts based on predicted user characteristics, so when users actually request content, the ads are already ready to be inserted immediately, eliminating processing delays while maintaining high relevance through cohort-based personalization.
Solution Approach 2:
The patent implements dynamics by enabling real-time cohort identification and ad manifest generation based on current user characteristics. The system dynamically assigns users to appropriate cohorts and retrieves pre-computed ad inserts on-demand, allowing personalized ad delivery without delay while adapting to changing user behavior patterns through continuous cohort updates.
Data Source
AI summary
The present disclosure provides methods and systems for real-time delivery of a target content in a streaming content. The method comprising receiving a request for streaming content from at least one user device [102]. The request comprises user segment information associated with the at least one user device. Next, the method comprises dynamically fetching at least one target content for the at least one user device [102] based on the user segment information. Next, the method comprises creating a customized virtual manifest by inserting the at least one target content in the requested streaming content in real-time. Next, the method comprises sending the customized virtual manifest to the at least one user device [102] based on the user segment information. Thereafter, the method comprises delivering the at least one target content in the requested streaming content to the at least one user device [102] based on the customized virtual manifest.